Reg Environ Change (2014) 14:157–166 DOI 10.1007/s10113-013-0475-x

ORIGINAL ARTICLE

Analysis of deforestation patterns in the central , Madagascar, between 1973 and 2010

Dietmar Zinner • Christian Wygoda • Leon Razafimanantsoa • Rodin Rasoloarison • Herizo T. Andrianandrasana • Jo¨rg U. Ganzhorn • Frank Torkler

Received: 17 September 2012 / Accepted: 3 May 2013 / Published online: 24 May 2013 Ó The Author(s) 2013. This article is published with open access at Springerlink.com

Abstract The central Menabe region still holds the largest Keywords Madagascar Deforestation Dry forest remnant of dry forest in western Madagascar. These forests Satellite image time series Remote sensing Political are home to high floral and faunal diversity including a crisis number of local and regional endemics. The forests of the central Menabe have been classified as conservation hot- spots. However, pressure on these forests is strong and Introduction deforestation continues on a large scale. To quantify recent forest loss, we used a series of satellite images (1973–2010) Madagascar, the world’s fourth largest island of approxi- for estimating annual deforestation rates. The overall rate mately 587,000 km2, is located off the south-eastern coast was 0.67 %, but it accelerated during certain periods to over of Africa. It is regarded as one of the major biodiversity 1.5 % with a maximum of 2.55 % per year between 2008 hotspots (Myers et al. 2000), because it harbours excep- and 2010. Not all areas within the forest block of the central tional biota with high overall diversity and levels of Menabe are affected similarly. Areas surrounding existing endemism (Goodman and Benstead 2005). At the same clearings show the highest losses of largely undisturbed time, its plant and species are among the most forest. If deforestation continues at the same rate as during endangered (Mittermeier et al. 2004; IUCN 2011). As in the last years, 50 % of the 1973 forest cover will be gone other parts of the tropics, habitat loss and fragmentation are within the next 11–37 years. among the most pervasive causes of biodiversity loss (Laurance et al. 2000; Geist and Lambin 2002; Fahrig 2003; Gardner et al. 2010; Irwin et al. 2010). The vast D. Zinner (&) majority of the terrestrial biota is forest dependent, but Cognitive Ethology Laboratory, Deutsches Primatenzentrum, deforestation has reached alarming proportions. Although Go¨ttingen, Germany e-mail: [email protected] historical estimates of national forest cover of Madagascar are difficult and somehow controversial (e.g. Dufils 2003), C. Wygoda F. Torkler the tendency of forest loss is obvious. Dufils (2003) Landscape Management and Nature Conservation, Hochschule reported deforestation rates for evergreen forest formations fu¨r nachhaltige Entwicklung Eberswalde (FH), Eberswalde, Germany of 102,000 ha per year in the period from 1950 to 1999. These estimates are similar to those of Green and Sussman L. Razafimanantsoa R. Rasoloarison (1990) for the eastern rain forests which had lost 50 % Behavioral Ecology and Sociobiology Unit, Deutsches between 1950 and 1985 with a rate of 111,000 ha per year. Primatenzentrum, Go¨ttingen, Germany Monitoring and analysing deforestation in Madagascar H. T. Andrianandrasana was focussed on the eastern humid forests (Green and Durrell Wildlife Conservation Trust, Antananarivo, Madagascar Sussman 1990; Nelson and Horning 1993; Dufils 2003; Bollen and Donati 2006;Va˚gen 2006). However, defor- J. U. Ganzhorn Biozentrum Grindel, Zoologisches Institut, estation rates in the western dry deciduous forests with Universita¨t Hamburg, Hamburg, Germany their comparable levels of endemism and biodiversity were 123 158 D. Zinner et al. similarly high or even higher (Smith 1997; Whitmore 2008); the only population of Madame Berthe’s mouse 2000; Harper et al. 2007; Scales 2011). It was estimated (Microcebus berthae), the smallest primate of the that over 97 % of Madagascar’s dry deciduous forests had world (Rasoloarison et al. 2000; Yoder et al. 2000); the flat- already disappeared by the year 2000 (WWF 2001) and tailed tortoise (Pyxis planicauda; Tidd et al. 2001); and a with them unique aspects of Malagasy biodiversity (Aglyptodactylus laticeps; Glaw et al. 1998; Glos et al. (Whitmore and Sayer 1992; Ganzhorn et al. 2001; Allnutt 2008). Sixteen other vertebrate species with extremely et al. 2008). This matches the situation of most tropical dry restricted ranges are found in the KAFC including the forests that have been neglected due to an emphasis on striped mongoose (Mungotictis decemlineata; Raz- humid tropical forests (Janzen 1988; Lerdau et al. 1991; afimanantsoa 2003) and the white-breasted mesite (Mesi- Bellefontaine et al. 1997). tornis variegata; Durbin et al. 2005; reviewed by Goodman Causes for forest removal in Madagascar vary regionally and Raselimanana 2008). Tree species diversity is notably (Gorenflo et al. 2011). However, the main direct cause of high with more than 200 species (Abraham et al. 1996; forest loss in western Madagascar is deforestation through Rakotonirina 1996). Current pressures in the region stem intentional burning to clear land for agriculture, locally primarily from an expanding human population which uses known as hatsake, either for subsistence or for the market slash-and-burn agriculture (for subsistence and cash crop), production (Scales 2011, 2012). With an expanding rural legal and illegal selective logging and animal trapping for population and increasing degradation of existing arable subsistence. There have been only five legal timber har- lands, the pressure on the remaining forest is extremely vesters in the region, four of which held harvesting lots high. Selective logging poses additional threats to forest defined by the service of Environment and Forests in the habitats (Whitehurst et al. 2009). Central Menabe region; all other activity was considered A priority-setting workshop concerning Madagascar in illegal. For instance, during the monitoring of the north- 1995 identified several sites of the dry deciduous forest eastern corner of the Kirindy Forest in 2002, the Waters and ecosystem in need of immediate biodiversity protection, Forests service seized 400 logs and destroyed 40 lemur traps. among them were the southern and the central Menabe In order to insure the survival of the endemic species and to regions (Ganzhorn et al. 1997; Hannah et al. 1998). In protect their habitat, an initiative was launched to improve 1997, a large contiguous block of the dry forest in the the legal protection of the KAFC and to create a protected southern Menabe became legally protected as the Kirindy- area, called ‘Aire Prote´ge´e Menabe Antimena’ (Ministe`re de Mitea National Park, one of the first national parks in the l’Environnement, des Eaux et Foreˆts 2006). western forest zone (De´cret no 97-1453, 18 December To further support the process towards creating a per- 1997 by the Ministe`re de l’Environnement, des Eaux et manent protected area in the central Menabe and to identify Foreˆts). The central Menabe region includes the area deforestation hotspots within the region, we performed a between the River in the south and the Tsiri- temporal and spatial analysis of forest loss in KAFC for the bihina River in the north. The northern part of the region, last 37 years (1973–2010) with special interest in the between Tomitsy and Tsiribihina rivers, still contains one deforestation rates during years of civil unrest, such as of the largest tracts of dry deciduous forests left in Mad- during the transition between the presidents Rasiraka and agascar, the Kirindy-Ambadira Forest Complex (KAFC). Ravalomanana from 2001/2002 and after Madagascar’s Prior to the establishment of Kirindy-Mitea National Park, unconstitutional change in government in 2009. the only protected area within the region was the Special Reserve Andranomena (approx. 7,000 ha), which is located south of the KAFC and south of the Tomitsy River. Study area However, it has already lost most of the local endemics because of forest degradation (Smith et al. 1997; Ganzhorn The study area is located in the dry deciduous forest of and Schmid 1998; Tidd et al. 2001). coastal western Madagascar between the Tomitsy The central Menabe has experienced a long history of (44°350E, 20°080S) and the Tsiribihina rivers (44°370E, forest conversion into agriculture, both for plantations and 19°440S) (Fig. 1). The Kirindy Forest constitutes the subsistence (Scales 2011), and further conversion of the southern part and the Ambadira Forest the northern part of remaining forests in the central Menabe poses serious threats the KAFC. In the west, the KAFC adjoins the coastal dunes to the survival of several endemic vertebrates. The global and marshes, and in the east, the forest borders savannah distribution of four vertebrate species is restricted to the grassland. The climate is characterized by pronounced forested areas between the Morondava and Tsiribihina riv- seasonality with little or no rain from April to November, ers, with the last remaining population of the giant jumping followed by a rainy season from December to March rat (Hypogeomys antimena), the largest living endemic (Ganzhorn 1995a; Sorg and Rohner 1996). The mean rodent of Madagascar (Sommer et al. 2002; Young et al. annual precipitation between 1906 and 1993 was 767 mm/year 123 Analysis of deforestation patterns 159

Fig. 1 Map of Madagascar showing the geographical position of the study area and the Kirindy-Ambadira Forest Complex (KAFC) in the central Menabe between Tomitsy and Tsiribihina rivers. CNFEREF = forest concession of the Centre National de Formation, d’Etudes et de Recherche en Environnement et Foresterie and the Kirindy Forest research station of the German Primate Center. Dark grey areas along the Tsiribihina River indicate lakes and marshes

(Sorg and Rohner 1996) with a tendency of increased canopy layer are deciduous; a herbaceous layer is absent variation over time (Ganzhorn 1995b). Mean annual rain- (except along dirt roads where grasses spread); flowering fall from 2000 to 2009 (including 3 years) at the often precedes the first flush of leaves; and trees show research station of the German Primate Center in the Ki- several adaptations to cope with prolonged water stress rindy Forest was 945 mm, also with increased variation (Sorg et al. 2003). For our classification, we followed over time (Kappeler and Dammhahn unpublished data). Rakotonirina’s (1996) description of forest structure in the Forest structure and phenology are not well studied in study area and defined forest as stands of trees with dbh most parts of the KAFC, with the exception of the conces- [10 cm, a height of [12 m and at least 75 % closed sion of the Centre de Formation Professionnelle Forestie`re, canopy during the rainy season (excluding secondary CFPF (now Centre National de Formation, d’Etudes et de regrowth, see below). Recherche en Environnement et Foresterie, CNFEREF), in We established 121 ground control points between 2001 the Kirindy Forest (Fig. 1). Ganzhorn and Sorg (1996) and 2010. Control points were either visited on the ground provide a detailed description of the site characteristics. or, in more remote areas, were evaluated when one of the Forest was described as a dense formation with a closed authors (HA) carried out an aerial survey in 2005. In an canopy at least during the rainy season when there is leaf area with at least a 50 m radius around each point, we cover and trees rarely exceed 20 m in height. Further assessed whether the vegetation corresponded to our forest information about forest structure and phenology for the definition. In zones with cleared forest, if not used for CNFEREF concession can be found in Rakotonirina (1996) agricultural purposes, the vegetation consists of secondary and Sorg and Rohner (1996). The KAFC is traversed by the formations, shrub and savannah grassland. Route Nationale 8 in north–south and in east–west directions by several straight 3–4-m-wide tracks cut through the forest Satellite data and analysis in the 1970s and 1980s for oil exploration. Remote sensing has been successfully applied in defores- tation monitoring in many studies in the tropics including Methods Madagascar (e.g. Nelson and Horning 1993; Vin˜a et al. 2004;Va˚gen 2006; Duveiller et al. 2008; Hansen et al. Forest classification 2008). With the opening of the Landsat 7 archive by the US Geological Service (USGS) in January 2009, a wide array The forest in our study area would be classified ‘Western of satellite scenes became available. To cover our complete dry forest’, and it shows a number of traits that are char- study period of more than 35 years, we had to use data acteristic for seasonally dry deciduous forests (Du Puy and from a variety of sensor systems, because no single system Moat 1996; Moat and Smith 2007): most trees of the was in operation long enough. Thus, four Landsat 7 scenes,

123 160 D. Zinner et al.

Fig. 2 Earth observation satellites and satellite scene dates used together with one Landsat 1, two Landsat 5 and two forest in the subsequent scene, we inferred forest presence ASTER scenes were the base for the establishment of the also for the 2001 scene. deforestation time series. Figure 2 shows scene dates used Finally, each forest/non-forest map was smoothed using for this analysis. All scenes have been resampled to the a39 3 mode filter in order to reduce noise of salt- extent and spatial resolution (pixel size 30 m 9 30 m) of and-pepper effects resulting from pixel-based NDVI clas- the 1992 Landsat scene. sification. These maps were aggregated into a single We aimed to select satellite scenes which were maxi- deforestation time series. Secondary regrowth is ignored: mally cloud free and which did not represent data from the pixels which were categorized as non-forest in one time rainy season. Most of our selected scenes were sampled in step, but were classified as (re)forested in one of the sub- May or June when the agricultural areas were already sequent time steps still were assigned as non-forest. This harvested and in comparison with forest almost vegetation accounts for the lower ecological value of second-growth free. Forests and harvested agricultural areas showed a forests (lower floral and faunal diversity), since the species significant difference in near infrared reflection and could composition of plant and animal communities often differs therefore be distinguished clearly. between secondary and old-growth forests (Chazdon 2003; Each Landsat scene was processed with an unsupervised Dunn 2004; Lugo and Helmer 2004; Gardner et al. 2007). classifier yielding a 25-class partition for each scene. These The ecological value of secondary forest in the central classes were combined with the derived Normalized Dif- Menabe has not been studied in detail (Sorg 2006), but by ference Vegetation Index to determine the mean NDVI comparing largely undisturbed and secondary forests in the value for each class (Ingram and Dawson 2005). In the region, it became obvious that lemur diversity and popu- used scenes, forests yield the highest NDVI values, and lation sizes are lower in secondary forests and endemic almost vegetation-free agricultural areas produced only species, such as Madame Berthe’s mouse lemur or the giant low NDVI values. A threshold value was established for jumping rat depend on dense old-growth dry forest and are each scene independently by visually comparing the not found in secondary regrowth (Smith 1997; Smith et al. resulting forest/non-forest map with the input satellite 1997; Ganzhorn et al. 1999; Sommer et al. 2002; Dam- scene texture of known forested and non-forested areas mhahn and Kappeler 2008). (based on several years of field experience of several of the authors in the region). The yielded threshold values range Temporal structuring from 0.2 to 0.4. Both ASTER scenes show areas of both good and bad We divided the total monitoring interval of 37 years plant conditions indicated by corresponding NDVI values: (1973–2010) into eight periods using the acquisition dates Smaller areas mainly in the north-east show high NDVI of the nine satellite images as separators. Since the eight values, most likely caused by local precipitations at the periods are not of equal lengths (Table 1), we normalized onset of the rain season in November 2008 (acquisition forest loss by using deforestation per year: forest loss date 03 December 2008), the rest show overall low values within period (ha)/length of period (d) * 365. due to long dryness. Therefore, each scene was divided into smaller portions—the assessment areas described in Spatial structuring Table 2 and shown on the result map (Fig. 5). These areas were assessed using the classifying/NDVI threshold We assumed that not all regions within the study area were approach described above. All used satellite scenes were similarly impacted by deforestation. Forest areas adjacent cloud free with the exception of the 2001 scene, where to villages and around already existing clearings probably 1.5 % of the area of interest was covered. Since opaque experience higher deforestation rates than others (Mertens clouds were only recorded above areas clearly classified as and Lambin 1997; Etter et al. 2006; Mon et al. 2012).

123 Analysis of deforestation patterns 161

Table 1 Dates and duration (days) of study periods Start End Duration

15 Jun 1973 20 May 1992 6,915 21 May 1992 02 May 2000 2,904 03 May 2000 03 Apr 2001 336 04 Apr 2001 22 Apr 2002 384 23 Apr 2002 20 Feb 2003 304 21 Feb 2003 04 Jun 2006 1,200 05 Jun 2006 02 Dec 2008 912 03 Dec 2008 06 May 2010 541 15 Jun 1973 06 May 2010 13,832

Table 2 List of assessment areas, area size and whether the areas contained at least one settlement

Name Id code Size (ha) Settlement Fig. 3 Absolute annual deforestation rates (ha) within KAFC. Lower dashed line: annual deforestation rate over the complete period of 1 7,112 Yes 37 years (1973–2010; 631.6 ha a-1 or 0.67 % a–1 of the 1973 forest Tsianaloka 2 3,269 Yes cover); upper dashed line: annual deforestation rate over the period of Ambadira 3 15,025 No the last 18 years (1992–2010; 872.7 ha a-1 or 1.0 % a-1 of the 1992 Salapeno 4 3,425 No forest cover). Numbers in columns refer to relative annual defores- tation rates (%) within each period Lambokely 5 2,328 Yes Masiposa 6 6,446 No Foreˆt d’Ankirisa 7 4,627 No North 8 8,798 Yes classifying forest-free areas as forest, we underestimated Corridor North 9 5,686 Yes the true proportion of deforestation making our classifica- Beroboka South 10 6,245 Yes tion a conservative and robust approach. The proportion of Corridor South 11 11,490 Yes deforestation tends to be even higher than our estimate. Kirindy Village 12 4,942 Yes In 1973, the remaining forest cover of the KAFC was approximately 94,743 ha. Within the subsequent 37 years, CNFEREF 13 18,037 No the forest cover was reduced by 23,353 ha (24.6 %) to only Marofandilia North 14 1,556 Yes 71,390 ha which means an average annual deforestation Marofandilia East 15 5,378 Yes rate of 631.6 ha or 0.67 % based on the 1973 forest cover. Total area 104,364 When taking into account only the period from 1992 to 2010, the annual rate was 872.7 ha or 1.00 % of the 1992 forest cover. Annual deforestation rates were relatively low Therefore, we partitioned the study area into 15 arbitrary between 1973 and 1992 (0.42 %) but exceeded 1.3 % geographical assessment areas for the statistical analysis. between 2000 and 2003 (Fig. 3). The rate subsequently Each area was named after a village, the concession holder dropped to less than 0.8 % in the period between 2003 and (e.g. CNFEREF) or the name was taken from topographical 2006. Recently, deforestation increased again above 1 % in maps (1:100.000, FTM, Institut National de Ge´ode´sie et de the period from 2006 to 2008 and up to 2.55 % between Cartographie 1956, reprint 1980) (Table 2). 2008 and 2010. Not all areas within the KAFC have been equally affected by forest loss. Deforestation rates were highest Results around certain settlements, namely Lambokely, Kirindy Village, Beroboka North and South and Marofandilia Our forest/non-forest classification in the final map was (Figs. 4, 5). verified at 121 ground control points. All 78 forest control The most dramatic forest loss was detected in the period points (100 %) and 88.4 % of the 43 non-forest points were from 2008 to 2010 as extensions of already existing agri- classified correctly (total classification user accuracy cultural clearings such as Lambokely, Beroboka North and 95.9 %), that is, we tended to classify more deforested South and along the northern limit of the KAFC at areas as forest than vice versa. By sometimes wrongly Tsimafana and Tsianaloka (Figs. 4, 5).

123 162 D. Zinner et al.

Fig. 4 Relative annual forest loss (%) of 15 areas within the KAFC in 8 study periods. Size of circles depict magnitudes of deforestation rates (larger circles correspond to higher deforestation rates). For higher rates exact figures are included

Table 3 Comparison of forest loss calculated by the present study (% per year) with data from Tidd et al. (2001) and Scales (2011) Period Forest loss a-1 % Study

1963–1984 0.40 Tidd et al. 1984–1993 1.10 Tidd et al. 1973–1992 0.4 Scales 1992–1999 0.4 Scales 1999–2005 1.0 Scales 1973–1992 0.42 This study 1992–2000 0.80 This study 2000–2010 1.24 This study

Discussion

The time series analysis of satellite images suggests rapid deforestation in the central Menabe. Our estimates of deforestation rates match those of Tidd et al. (2001) for the region between the Tomitsy River and the Tsiribihina River and those of Scales (2011). Although their study periods are not directly comparable to ours and their forest definition seems to be broader than the one we applied, the magnitude of annual deforestation rates corresponds to our results (Table 3), with a general acceleration over the years despite the fact that forest clearance for agriculture has been illegal since 1987 (De´cret no 87-143, 20 April 1987 by the Ministe`re de l’Environnement, des Eaux et Foreˆts). Deforestation of the area was already documented by Smith (1997) and Smith et al. (1997). Deforestation in Menabe is regarded as an almost irreversible process, Fig. 5 Map of forest loss in the KAFC 1973–2010. Grey colour indicates remaining forest cover in 2010. Study area partition is because these forests have poor regenerative power due to overlaid the scarcity or lack of secondary colonizing species and 123 Analysis of deforestation patterns 163 disturbed areas are rapidly invaded by alien plant species activities in the central Menabe have been extremely high (Genini 1996; Lowry II 1997). Also, secondary regrowths during times when central governmental control was low, a are unsuitable habitats for the endemic forest species (e.g. temporal coincidence also reported by Va˚gen (2006) for Madame Berthe’s mouse lemur, giant jumping rat; Sommer deforestation in parts of the eastern highlands of Mada- et al. 2002; Dammhahn and Kappeler 2008). Thus, the gascar. In the short period between 2000 and 2003, when a value of such secondary habitats for preserving the forest political conflict almost turned into a civil war, deforesta- endemics is highly questionable. tion increased from less than 500 ha per year (0.5 %) in the \Reasons for deforestation are manifold as in other parts of period from 1973 to 2000 to more than 1,100 ha (1.4 %) in Madagascar (Brooks et al. 2009; Gorenflo et al. 2011). In the the following three years. In the post-crisis period from case of central Menabe, forest clearance for agriculture is the 2003 to 2008, intensive conservation activities were laun- most important proximate factor (Smith 1997; Smith et al. ched in the central Menabe with a series of talks with local 1997). This is particularly obvious in the vicinity of already authorities and villagers, confiscation of illegally logged existing slash-and-burn areas such as the hatsake of Lamb- wood, surveys and monitoring activities. During this per- okely, Kirindy Village, Beroboka North and South and iod, the deforestation rate dropped to 706 ha per year Marofandilia. However, as Scales (2011, 2012) pointed out, (0.9 %). In 2009 and 2010, with a new political crisis different social, political and economic factors drive defor- associated with considerable civil unrest and reduced estation in the central Menabe as in other parts of south- governmental control, the deforestation rate in the central western Madagascar (Casse et al. 2004). For instance, in Menabe increased to a new record high of more than Tsimafana (zone 1), forest is converted in large quantities into 1,890 ha per year (2.55 %). Although we cannot provide maize fields for market production or in the north-eastern any statistical evidence for a relationship between political areas of Menabe (zones 4 Salapeno, 6 Masiposa and 7 An- instability with low legal enforcement and the increase in kirisa), the extension of cattle pasture seems to be the main conversion of forest into agriculture, the temporal coinci- cause for habitat clearing, as local people are predominantly dences are notable. An elevated deforestation rate during cattle breeders and only marginally practise farming. These times of political instability, in particular on the subsis- areas are of particular interest because deforestation has taken tence level, seems to be a widespread phenomenon (Dea- place here within the last years without being noticed by the con 1994; Didia 1997;Va˚gen 2006; Randriamalala and Liu local authorities. Large concessions and plantations (e.g. the 2010; Freudenberger 2010; Innes 2010; World Bank 2010). sisal plantation around Beroboka) with subsequent immigra- In the interest of conservation of natural resources and tion of non-autochthonous people as workers into the area had biodiversity in the Central Menabe region, all relevant an over-proportional impact on deforestation and the stakeholders such as the Fanamby NGO, Durrell Wildlife demography of the region (Scales 2011). Conservation Trust, Madagascar National Parks, CNFE- The extension of already existing clearings not only REF, WWF, OPCI (l’Organisation Publique de Coope´ra- reduces forest cover but can result also in forest fragmen- tion Intercommunale) and government offices have united tation. The danger of fragmentation becomes particularly to form the ‘Platform for Conservation of Biodiversity and obvious in the Corridor South area (area partition 11) Development in Central Menabe’. The ‘platform’ agreed where further forest loss will immediately lead to the that the weak or even non-existing law enforcement is one separation of the southern Kirindy and the northern Am- of the key issues in relation to deforestation in Menabe. badira forests. The cleared areas would constitute a dis- The regional direction of environment and forests (DREF) persal barrier for forest-dependent species, resulting in a has recently set up the service of investigation and control cut-off of gene flow between the Kirindy and Ambadira (SRIC) which manages infractions and law. The main parts of the KAFC. There is already evidence for a cut-off. concern of this service is to investigate and follow-up A rapid lemur survey in 2007 (Kappeler et al. unpublished threats and pressures in the forest. However, apparently the data) revealed that the corridor is only functional for service is lacking funds and personnel to implement field Microcebus murinus and to a lesser degree for Mirza co- visits and control or to work closely with local communi- quereli and Phaner furcifer. ties to assist through participatory ecological monitoring. In addition to the results of the deforestation identified by One option to reduce deforestation in Menabe would be satellite images, selective logging is another factor that has a technical and financial support of the SRIC. negative impact on the Menabe forests. In contrast to slash- and-burn, which is obvious in the satellite data, selective log- ging also degrades the forest, but is difficult to measure with Future perspectives satellite images. Reliable data on logging are not available. It might be that deforestation in KAFC was accelerated Assuming that deforestation will continue at the same by an additional factor. It is noticeable that deforestation amount (631.2 ha a-1) as over the whole study period from 123 164 D. Zinner et al.

References

Abraham JP, Rakotonirina B, Randrianasolo M, Ganzhorn JU, Jeannoda V, Leigh EG (1996) Tree diversity on small plots in Madagascar: a preliminary review. Rev Ecol (Terre Vie) 51:93–116 Allnutt TF, Ferrier S, Manion G, Powell GVN, Ricketts TH, Fisher BL, Harper GJ, Irwin ME, Kremen C, Labat JN, Lees DC, Pearce TA, Rakotondrainibe F (2008) A method for quantifying biodiversity loss and its application to a 50-year record of deforestation across Madagascar. Conserv Lett 1:173–181. doi: 10.1111/j.1755-263X.2008.00027.x Bellefontaine R, Gaston A, Petrucci Y (1997) Ame´nagement des foreˆts naturelles des zones tropicales se`ches. Cahier Conserva- tion FAO 32. FAO, Rome Bollen A, Donati G (2006) Conservation status of the littoral forest of south-eastern Madagascar: a review. Oryx 40:57–66. doi: 10.1017/S0030605306000111 Brooks CP, Holmes C, Kramer K, Barnett B, Keitt TH (2009) The role of demography and markets in determining deforestation rates near Ranomafana National Park, Madagascar. PLoS ONE Fig. 6 Reduction in forest cover in the KAFC and projected change 4:e5783. doi:10.1371%2Fjournal.pone.0005783 under four deforestation scenarios: an annual deforestation (1) similar Casse T, Milhøj A, Ranaivoson S, Randriamanarivo J (2004) Causes to the overall annual amount in the period from 1973 to 2010; (2) of deforestation in southwestern Madagascar: what do we know? similar to the annual amount in the period from 1992 to 2010; (3) Forest Pol Econ 6:33–48. doi:10.1016/S1389-9341(02)00084-9 similar to the annual amount within the last 10 years; and (4) similar Chazdon RL (2003) Tropical forest recovery: legacies of human to the annual amount in the period from 2008 to 2010. Forest cover in impact and natural disturbances. Perspect Plant Ecol Evol 1973 was set at 100 % Systemat 6:51–71. doi:10.1078/1433-8319-00042 Dammhahn M, Kappeler PM (2008) Small-scale coexistence of two mouse lemur species (Microcebus berthae and M. murinus) 1973 to 2010, the KAFC will have lost 50 % of the 1973 within a homogeneous competitive environment. Oecologia forest cover by 2049 (Fig. 6). In case of deforestation 157:473–483. doi:10.1007/s00442-008-1079-x continues with an average rate similar to that over the Deacon RT (1994) Deforestation and the rule of law in a cross-section -1 of countries. Land Econ 70:414–430 period from 1992 to 2010 (713.9 ha a ), the forest cover Didia DO (1997) Democracy, political instability and tropical will be reduced to only 50 % by 2043. With an annual deforestation. Global Environ Change 7:63–76. doi:10.1016/ deforestation of 885.2 ha a-1 similar to the rate within the S0959-3780(96)00024-6 last 10 years, 50 % of the 1973 forest cover will be gone by Du Puy D, Moat J (1996) A refined classification of the primary vegetation of Madagascar based on the underlying geology: 2037. However, if deforestation continues with an average using GIS to map its distribution and to assess its conservation annual deforestation as between 2008 and 2010 status. In: Lourenc¸o WR (ed) Proceedings of the International (1,820.4 ha a-1), but with no further acceleration, the Symposium on the ‘Biogeography de Madagascar’, Paris, September 1995. Editions de l’ORSTOM, Paris, pp 205–218 forest cover will be reduced to 50 % by 2023. Such an Dufils J (2003) Remaining forest cover. In: Goodman SM, Benstead annual deforestation will lead to a complete loss of forest JP (eds) The natural history of Madagascar. The University of by 2,049 and with it most, if not all, endemic species. Chicago Press, Chicago, pp 88–96 These projections show alarming trends and are cause for Dunn RR (2004) Recovery of faunal communities during tropical serious concern and should lead to an acceleration of forest regeneration. Conservat Biol 18:302–309. doi:10.1111/j. 1523-1739.2004.00151.x conservation activities. Durbin J, Ratsimbazafy J, Volahy A, Kappeler PM, Zinner D, Rasoloarison R, Razafimantsoa L, Glos J, Andrianjanahary T Acknowledgments We would like to thank the Global Conserva- (2005) Etudes sur la Biodiversite´ de Menabe Central. Unpub- tion Program (GCP) of Conservation International for financial sup- lished Report. Durrell Wildlife Trust and Deutsches Primaten- port, the Aviation sans Frontie`re Belge for their help with the aerial zentrum, Antananarivo, Madagascar surveys, the local forest committees, Anselme Toto Volahy and Duveiller G, Defourny P, Descle´e B, Mayaux P (2008) Deforestation Francisco Rakotombololona for their assistance during the ground in Central Africa: estimates at regional, national and landscape truthing process. We thank the Direction Re´gionale de l’Environne- levels by advanced processing of systematically-distributed ment et des Foreˆts de Menabe and the Madagascar National Parks for Landsat extracts. Remote Sens Environ 112:1969–1981. doi: their cooperation. Thanks to Tanja Haus, Gisela Fickenscher, Melanie 10.1016/j.rse.2007.07.026 Dammhahn and Moritz Rahlfs for their comments on an earlier ver- Etter A, McAlpine C, Wilson K, Phinn S, Possingham H (2006) sion of the paper. Regional patterns of agricultural land use and deforestation in Colombia. Agric Ecosyst Environ 114:369–386. doi:10.1016/j. Open Access This article is distributed under the terms of the agee.2005.11.013 Creative Commons Attribution License which permits any use, dis- Fahrig L (2003) Effects of habitat fragmentation on biodiversity. tribution, and reproduction in any medium, provided the original Annu Rev Ecol Evol Syst 34:487–515. doi:10.1146/annurev. author(s) and the source are credited. ecolsys.34.011802.132419

123 Analysis of deforestation patterns 165

Freudenberger K (2010) Paradise lost? Lessons from 25 Years of Hannah L, Rakotosamimanana B, Ganzhorn JU, Mittermeier RA, USAID Environment Programs in Madagascar. International Olivieri S, Iyer L, Rajaobelina S, Hough J, Andriamialisoa F, Resources Group. United States Agency for International Bowles I, Tilkin G (1998) Participatory planning, scientific Development (USAID), Washington, USA http://www.usaid. priorities, and landscape conservation in Madagascar. Environ gov/locations/sub-saharan_africa/countries/madagascar/ Conserv 25:30–36. doi:10.1017/S0376892998000071 paradise_lost_25years_env_programs.pdf Hansen MC, Stehman SV, Potapov PV, Loveland TR, Townshend Ganzhorn JU (1995a) Low-level forest disturbance effects on primary JRG, DeFries RS, Pittman KW, Arunarwati B, Stolle F, production, leaf chemistry, and lemur populations. Ecology Steininger MK, Carroll M, DiMiceli C (2008) Humid tropical 76:2084–2096. doi:10.2307/1941683 forest clearing from 2000 to 2005 quantified by using multitem- Ganzhorn JU (1995b) Cyclons over Madagascar: fate or fortune? poral and multiresolution remotely sensed data. P Natl Acad Sci Ambio 24:124–125. doi:10.2307/4314308 USA 105:9439–9444. doi:10.1073/pnas.0804042105 Ganzhorn JU, Schmid J (1998) Different population dynamics of Harper G, Steininger M, Tucker C, Juhn D, Hawkins F (2007) Fifty Microcebus murinus in primary and secondary deciduous dry years of deforestation and forest fragmentation in Madagascar. forests of Madagascar. Int J Primatol 19:785–796. doi:10.1023/ Environ Conserv 34:1–9. doi:10.1017/S0376892907004262 A:1020337211827 Ingram JC, Dawson TP (2005) Inter-annual analysis of deforestation Ganzhorn JU, Sorg JP (Eds) (1996) Ecology and economy of a hotspots in Madagascar from high temporal resolution satellite tropical dry forest in Madagascar. Primate Report 46:1–382 observations. Int J Remote Sens 26:1447–1461. doi:10.1080/ Ganzhorn JU, Rakotosamimanana B, Hannah L, Hough J, Iyer L, 01431160412331291189 Olivieri S, Rajaobelina S, Rodstrom C, Tilkin G (1997) Priorities Innes JL (2010) Madagascar rosewood, illegal logging and the for biodiversity conservation in Madagascar. Primate Report tropical timber trade. Madagascar Conserv Develop 5:6–10 48:1–81 Irwin MT, Wright PC, Birkinshaw C, Fisher B, Gardner CJ, Glos J, Ganzhorn JU, Fietz J, Rakotovao E, Schwab D, Zinner D (1999) Goodman SM, Loiselle P, Rabeson P, Raharison JL, Raherilalao and the regeneration of dry deciduous forest in MJ, Rakotondravony D, Raselimanana A, Ratsimbazafy J, Madagascar. Conserv Biol 13:794–804. doi:10.1046/j. Sparks J, Wilme´ L, Ganzhorn JU (2010) Patterns of species 1523-1739.1999.98245.x change in anthropogenically disturbed habitats of Madagascar. Ganzhorn JU, Lowry PP, Schatz GE, Sommer S (2001) The Biol Conserv 143:2351–2362. doi:10.1016/j.biocon.2010.01.023 biodiversity of Madagascar: one of the world’s hottest hotspots IUCN (2011) IUCN Red List of threatened species. Version 2011.2. on its way out. Oryx 35:346–348. doi:10.1046/j.1365-3008. www.iucnredlist.org. (Downloaded on 11 April 2012) 2001.00201.x Janzen DH (1988) Tropical dry forests: the most endangered major Gardner TA, Ribeiro-Ju´nior MA, Barlow J, Sauer Avila-Pires TC, tropical ecosystem. In: Wilson EO (ed) Biodiversity. National Hoogmoed MS, Peres CA (2007) The value of primary, Academy Press, Washington, pp 130–137 secondary, and plantation forests for a Neotropical herpetofauna. Laurance WF, Delamonica P, Laurance SG, Vasconcelos HL, Conserv Biol 21:775–787. doi:10.1111/j.1523-1739. Lovejoy TE (2000) Conservation: rainforest fragmentation kills 2007.00659.x big trees. Nature 404:836. doi:10.1038/35009032 Gardner TA, Barlow J, Sodhi NS, Peres CA (2010) A multi-region Lerdau M, Whitbeck J, Holbrook NM (1991) Tropical deciduous assessment of tropical forest biodiversity in a human-modified forest: death of a biome. Trends Ecol Evol 6:201–233. doi: world. Biol Conservat 143:2293–2300. doi:10.1016/j.biocon. 10.1016/0169-5347(91)90019-T 2010.05.017 Lowry PP II (1997) The classification of natural and anthropogenic Geist HJ, Lambin EF (2002) Proximate causes and underlying driving vegetation in Madagascar. In: Patterson BD, Goodman SM (eds) forces of tropical deforestation. Bioscience 52:143–150. doi: Natural change and human impact in Madagascar. Smithsonian 10.1641/0006-3568(2002)052[0143:PCAUDF]2.0.CO;2 Institution, Washington, pp 93–123 Genini M (1996) Deforestation. Primate Report 46:49–55 Lugo AE, Helmer E (2004) Emerging forests on abandoned land: Glaw F, Vences M, Bo¨hme W (1998) Systematic revision of the Puerto Rico’s new forests. Forest Ecol Manag 190:145–161. doi: genus Aglyptodactylus Boulenger, 1919 (Amphibia: Ranidae), 10.1016/j.foreco.2003.09.012 and analysis of its phylogenetic relationships to other Madaga- Mertens B, Lambin EF (1997) Spatial modelling of deforestation in scan ranid genera (Tomopterna, Boophis, Mantidactylus, and southern Cameroon: spatial disaggregation of diverse deforesta- Mantella). J Zool Syst Evol Res 36:17–37. doi:10.1111/j. tion processes. Appl Geogr 17:143–162. doi:10.1016/S0143- 1439-0469.1998.tb00775.x 6228(97)00032-5 Glos J, Dausmann KH, Linsenmair KE (2008) Modeling the habitat Ministe`re de l’Environnement, des Eaux et Foreˆts (2006) Arreˆte´ no use of Aglyptodactylus laticeps, an endangered dry-forest frog 4532/2006-MINENVEF portant protection temporaire de l’Aire from Western Madagascar. Monografie del Museo Regionale di Prote´ge´e en cre´ation de´nomme´e «Menabe central» Districts de Scienze Naturali di Torino 45:125–142 Morondava, Belo sur Tsiribihina et Mahabo Re´gion Menabe Goodman SM, Benstead JP (2005) Updated estimates of biotic Mittermeier RA, Langrand O, Lowry PP II, Schatz G, Gerlach J, diversity and endemism for Madagascar. Oryx 39:73–77. doi: Goodman SM, Steininger M, Hawkins F, Raminosoa N, 10.1017/S0030605305000128 Ramilijaona O, Andriamaro L, Randrianasolo H, Rabarison H, Goodman SM, Raselimanana A (2008) Exploration et connaissance Rakotobe ZL (2004) Madagascar and the Indian Ocean Islands. biologique des diffe´rents sites inventorie´s. Malagasy Nat In: Lamoreux J, da Fonseca GAB, Mittermeier RA, Robles Gil P, 1:33–45 Hoffman M, Pilgrim J, Brooks T, Goetsch Mittermeier C (eds) Gorenflo LJ, Corson C, Chomitz KM, Harper G, Honza´kM,O¨ zler B Hotspots revisited earth’s biologically richest and most endan- (2011) Exploring the association between people and defores- gered ecoregions. CEMEX, Mexico City, pp 138–144 tation in Madagascar. In: Cincotta RP, Gorenflo LJ (eds) Human Moat J, Smith P (eds) (2007) Atlas of the Vegetation of Madagascar. population: its influences on biological diversity. Springer, Royal Botanic Gardens, Kew. http://www.vegmad.org/ Berlin, pp 197–221 Mon MS, Mizoue N, Htun NZ, Kajisa T, Yoshida S (2012) Factors Green GM, Sussman RW (1990) Deforestation history of the eastern affecting deforestation and forest degradation in selectively rain forests of Madagascar from satellite images. Science logged production forest: a case study in Myanmar. Forest Ecol 248:212–215. doi:10.1126/science.248.4952.212 Manag 267:190–198. doi:10.1016/j.foreco.2011.11.036

123 166 D. Zinner et al.

Myers N, Mittermeier RA, Goetsch Mittermeier C, da Fonseca GAB, Sorg JP, Rohner U (1996) Climate and tree phenology of the dry Kent J (2000) Biodiversity hotspots for conservation priorities. deciduous forest of the Kirindy Forest. Primate Report 46:57–80 Nature 403:853–858. doi:10.1038/35002501 Sorg JP, Ganzhorn JU, Kappeler PM (2003) Forestry and research in Nelson R, Horning N (1993) AVHRR-LAC estimates of forest area in the Kirindy Forest/Centre de Formation Professionnelle Forest- Madagascar, 1990. Int J Remote Sens 14:1463–1475. doi: ie`re. In: Goodman SM, Benstead JP (eds) The natural history of 10.1080/01431169308953981 Madagascar. The University of Chicago Press, Chicago, Rakotonirina (1996) Composition and structure of a dry forest on pp 1512–1519 sandy soils near Morondava. Primate Report 46:81–87 Tidd ST, Pinder JE III, Ferguson GW (2001) Deforestation and Randriamalala H, Liu Z (2010) Rosewood of Madagascar: between habitat loss for the Malagasy flat-tailed tortoise from 1963 democracy and conservation. Madagascar Conserv Develop through 1993. Chelonian Conserv Biol 4:59–65 5:11–22 Va˚gen TG (2006) Remote sensing of complex land use change Rasoloarison RM, Goodman SM, Ganzhorn JU (2000) Taxonomic trajectories: a case study from the highlands of Madagascar. Agr revision of mouse lemurs (Microcebus) in the western portions Ecosyst Environ 115:219–228. doi:10.1016/j.agee.2006.01.007 of Madagascar. Int J Primatol 21:963–1019. doi:10.1023/ Vin˜a A, Echavarria FR, Rundquist DC (2004) Satellite change A:1005511129475 detection analysis of deforestation rates and patterns along the Razafimanantsoa L (2003) Mungotictis decemlineata, narrow-striped Colombia: Ecuador border. Ambio 33:118–125. doi:10.1579/ mongoose. In: Goodman SM, Benstead JP (eds) The natural 0044-7447-33.3.118 history of Madagascar. The University of Chicago Press, Whitehurst AS, Sexton JO, Dollar L (2009) Land cover change in Chicago, pp 1357–1360 western Madagascar’s dry deciduous forests: a comparison of Scales I (2011) Farming at the forest frontier: land use and landscape forest changes in and around Kirindy Mite National Park. Oryx change in Western Madagascar, 1896–2005. Environ Hist 43:275–283. doi:10.1017/S0030605309001756 17:499–524. doi:10.3197/096734011X13150366551481 Whitmore TC (2000) Madagascar deforestation rate during. In: Scales I (2012) Lost in translation: conflicting views of deforestation, Lourenc¸o WR, Goodman SM (eds) Diversite´ et Endemisme a` land use and identity in western Madagascar. Geogr J 178:67–79. Madagascar. Me´moires de la Socie´te´ de Bioge´ographie, Paris, doi:10.1111/j.1475-4959.2011.00432.x p 125 Smith AP (1997) Deforestation, fragmentation, and reserve design in Whitmore TC, Sayer JA (eds) (1992) Tropical deforestation and western Madagascar. In: Laurance WF, Bierregaard OW (eds) species extinction. Chapman & Hall, New York Tropical forest remnants: ecology, management and conserva- World Bank (2010) Governance and development effectiveness tion of fragmented communities. The University of Chicago review. A political economy analysis of governance in Mada- Press, Chicago, pp 415–441 gascar. World Bank, Report No. 54277-MG Smith AP, Horning N, Moore D (1997) Regional biodiversity WWF (2001) Madagascar Dry Deciduous Forests. http://www. planning and lemur conservation with GIS in western Madagas- worldwildlife.org/wildworld/profiles/terrestrial/at/ car. Conserv Biol 11:498–512. doi:10.1046/j.1523-1739. at0202_full.html. Accessed 03 December 2010 1997.95283.x Yoder AD, Rasoloarison RM, Goodman SM, Irwin JA, Atsalis S, Sommer S, Toto Volahy A, Seal U (2002) A population and habitat Ravosa MJ, Ganzhorn JU (2000) Remarkable species diversity in viability assessment for the highly endangered giant jumping rat Malagasy mouse lemurs (Primates, Microcebus). Proc Natl Acad (Hypogeomys antimena), the largest extant endemic rodent of Sci USA 97:11325–11330. doi:10.1073/pnas.200121897 Madagascar. Anim Conserv 5:263–273. doi:10.1017/S136 Young RP, Volahy AT, Bourou R, Lewis R, Durbin J, Wright TJ, 7943002004018 Hounsome TD, Fa JE (2008) A baseline estimate of population Sorg JP (2006) Orientations nouvelles de la recherche dans la zone size for monitoring the Endangered Madagascar giant jumping des foreˆts denses se`ches a` Madagascar. In: Schwitzer C, Brandt rat Hypogeomys antimena. Oryx 42:584–591. doi:10.1017/ S, Ackermand D, Ramilijaona O, Randria G, Ganzhorn JU (eds) S0030605307000816 Proceedings of the German-Malagasy research cooperation in life and earth sciences. Concept Verlag, Berlin, pp 67–71

123